Documente Academic
Documente Profesional
Documente Cultură
ISSN No:-2456-2165
However the reviewed techniques did not employ any A. Data Collection
machine learning algorithm to train, classify and enable the Our video surveillance system would monitor the
system detect and even predict any vandalism and other environment (pipeline installation route) and capture
illegal activities before they occur. Our proposed system various forms of activities, behavior, or other changing
intends to address that challenge as the emphasis is on information, usually, of people from a distance by means of
prevention by developing a system that not only detects but sensing device. This is the first step towards prevention,
also predicts possible occurrence. reaction, detection, intervention. The data captured is
transmitted in real time to a base facility.
B. Pre-processing
It is critical to carry out the data processing on the Fig 2:- Block Diagram of Deep Surveillance Human
collected surveillance feed as the quality of the surveillance Activity Recognition System.
might be affected by weather, poor lighting signal
interference and so on. Some of the preprocessing activities IV. CONCLUSION
includes, compression, de-noising, segmentation,
equalization etc. The impacts of this research are numerous. Some of
the benefits are; prevention of pipeline vandalism,
C. Feature Extraction reduction in financial losses, alleviation of environmental
Feature extraction is carried out on the preprocessed degradation, security and safety, increased oil revenue
surveillance feed. First of all the background subtraction amongst. By adopting state-of-the-art surveillance devices
operation is done on the surveillance, then the feature of and a robust machine learning algorithm, the proposed
interest is extracted for clustering and classification. system would be able to completely solve the problem of
pipeline vandalism which poses a major threat to the
D. Clustering and Activity Classification operations and optimal yield of our oil and gas industry.
The Support-Vector-Machine-(SVM) and K-Nearest-
Neighbor (KNN) are the machine learning tools that would REFERENCES
be employed to cluster, classify and train the system based
on the collected surveillance dataset on different activities, [1]. T. AL-Kadi, Z. AL-Tuwaijri, and A. AL- Omran,
movements and positions etc. Unlike conventional “Wireless Sensor Networks for Leakage Detection in
approaches this technique would enable the system make Underground Pipelines: A Survey Paper,” 5th Int.
smart and intelligent inferences and predictions on possible Symp. Appl. Ad hoc Sens. Networks, vol. 21, no. 0,
threats even before they occur. Equation 1 shows the pp. 491–498, 2013.
mathematical model of KNN prediction. [2]. A. F. Colombo, P. Lee, and B. W. Karney, “A
selective literature review of transient-based leak
1
𝑌 = 𝐾 ∑𝑘𝑖=1 𝑦𝑖 (1) detection methods,” J. Hydro-Environment Res., vol.
2, no. 4, pp. 212–227, 2009.
[3]. J. Sun and J. Wen, “Research on Monitoring and Pre-
Here, yi is define as the ith case of the examples
warning System for Security of Pipelines Based on
sample and y denoted to be the prediction (outcome) of the
Multi-Seismic Sensors,” Int. Conf. Electron. Meas.
query point.
Instruments, pp. 1–5, 2009.
[4]. W. Liang, J. Hu, L. Zhang, C. Guo, and W. Lin,
E. Activity Recognition
“Assessing and classifying risk of pipeline third-party
Our system would track individual as they pass
interference based on fault tree and SOM,” Eng. Appl.
through the field of vision of the camera, and with machine
Artif. Intell., vol. 25, no. 3, pp. 594–608, Apr. 2012.
learning algorithms to classify the motion and activities of
[5]. I. Jawhar, N. Mohamed, and K. Shuaib, “A
every passerby that walks a stipulated radius to our pipeline
framework for pipeline infrastructure monitoring
infrastructure. The tracking is accomplished through the
using wireless sensor networks,” in Wireless
development of a position and velocity path characteristic.
Telecommunications Symposium, 2007. WTS 2007,
With this information, our system would report any threat
2007, pp. 1–7.
or suspected threat to the attention of human security
personnel.